Lymphoma is a well known complication of HIV infection. Previuos studies have identified several mechanisms of B cell activation in patients with HIV infection which may lead to lymphoma: 1) increased numbers of Epstein-Barr virus infected B cells; 2) T cell dependent polyclonal B cell activation induced by HIV; 3) antigen-specific B cell activation by HIV; 4) increased IL-6 production induced by HIV. We have recently observed that 8 of 55 patients (14.5%) with AIDS or severe ARC entered onto three long-term phase I trials of azidothymidine (AZT) or AZT-based regimens at the NCI between 1985 and 1987 developed a high-grade non- Hodgkin's lymphoma (NHL), B cell type, after initiating antiretroviral therapy. The NHL occurred a median of 23.6 months after initiating antiretroviral treatment. The estimated probability of developing lymphoma by 24 months of therapy was 12%, increasing to 29% after 36 months. The patients who developed NHL had a median of 6 CD4 cells/mm3 at the time of NHL diagnosis. There was a statistically significant difference in the rate of development of NHL during the time the patients had 50 CD4 cells compared to the time they had 50 CD4 cells, independen of the time they received antiretroviral therapy. Patients with symptomatic HIV infection who survive for prolonged periods while severely immunosuppressed have a relatively high probability of developing lymphoma. We are currently examining a number of parameters to determine which factors are predictive for the subsequent development of NHL. As improved therapies for the treatment of HIV and its complications result in prolonged survival, NHL may become an increasingly significant problem. In collaboration with Dwight Kaufman, M.D. in the Radiation Oncology Branch, we are conducting a trial of combination chemotherapy with AZT and GM-CSF. Preliminary results suggest that some patients may respond to this regimen but that toxicity can be a problem.